Background Rising levels of obesity require interventions that support people in long-term weight loss. The Football Fans in Training (FFIT) programme uses loyalty to football teams to engage men in weight loss. In 2011/12, a randomised controlled trial (RCT) found that the FFIT programme was effective in helping men lose weight up to 12 months. Objectives To investigate the long-term weight, and other physical, behavioural and psychological outcomes up to 3.5 years after the start of the RCT; the predictors, mediators and men’s qualitative experiences of long-term weight loss; cost-effectiveness; and the potential for long-term follow-up via men’s medical records. Design A mixed-methods, longitudinal cohort study. Setting Thirteen professional Scottish football clubs from the RCT and 16 additional Scottish football clubs that delivered the FFIT programme in 2015/16. Participants A total of 665 men who were aged 35–65 years at the RCT baseline measures and who consented to follow-up after the RCT (intervention group, n = 316; comparison group, n = 349), and 511 men who took part in the 2015/16 deliveries of the FFIT programme. Interventions None as part of this study. Main outcome measures Objectively measured weight change from the RCT baseline to 3.5 years. Results In total, 488 out of 665 men (73.4%) attended 3.5-year measurements. Participants in the FFIT follow-up intervention group sustained a mean weight loss from baseline of 2.90 kg [95% confidence interval (CI) 1.78 to 4.02 kg; p < 0.001], and 32.2% (75/233) weighed ≥ 5% less than at baseline. Participants in the FFIT follow-up comparison group (who participated in routine deliveries of the FFIT programme after the RCT) lost a mean of 2.71 kg (95% CI 1.65 to 3.77 kg; p < 0.001), and 31.8% (81/255) achieved ≥ 5% weight loss. Both groups showed long-term improvements in body mass index, waist circumference, percentage body fat, blood pressure, self-reported physical activity (PA) (including walking), the consumption of fatty and sugary foods, fruit and vegetables and alcohol, portion sizes, self-esteem, positive and negative affect, and physical and mental health-related quality of life (HRQoL). Mediators included self-reported PA (including walking) and sitting time, the consumption of fatty and sugary foods and fruit and vegetables, portion sizes, self-esteem, positive affect, physical HRQoL, self-monitoring of weight, autonomous regulation, internal locus of control, perceived competence, and relatedness to other FFIT programme participants and family members. In qualitative interviews, men described continuing to self-monitor weight and PA. Many felt that PA was important for weight control, and walking remained popular; most were still aware of portion sizes and tried to eat fewer snacks. The FFIT programme was associated with an incremental cost-effectiveness of £10,700–15,300 per quality-adjusted life-year (QALY) gained at 3.5 years, and around £2000 per QALY gained in the lifetime analysis. Medical record linkage provided rich information about the clinical health outcomes of the FFIT RCT participants, and 90% of men (459/511) who took part in the 2015/16 FFIT programme gave permission for future linkage. Conclusions Participation in the FFIT programme under both research (during the FFIT RCT) and routine (after the FFIT RCT) delivery conditions led to significant long-term weight loss. Further research should investigate (1) how to design programmes to improve long-term weight loss maintenance, (2) longer-term follow-up of FFIT RCT participants and (3) very long-term follow-up via medical record linkage. Trial registration Current Controlled Trials ISRCTN32677491. Funding This project was funded by the National Institute for Health Research (NIHR) Public Health Research programme and will be published in full in Public Health Research; Vol. 6, No. 9. See the NIHR Journals Library website for further project information. The Scottish Executive Health Department Chief Scientist Office (CSO) funded the feasibility pilot that preceded the FFIT RCT (CZG/2/504). The Medical Research Council (MRC) funded Kate Hunt and additional developmental research through the MRC/CSO Social and Public Health Sciences Unit Gender and Health programme (5TK50/25605200-68094).